Skip to content

Methods and classes to support Biosynthetic Gene Cluster analysis

License

Notifications You must be signed in to change notification settings

jorgecnavarrom/BGClib

Repository files navigation

BGClib

This is the official repository of the Biosynthetic Gene Cluster Library, a Python library with classes and methods for manipulation and analysis of BGCs.

For more information on BGClib, see here.

⚠️ This code is under heavy development!

BGC Toolkit

This is a tool that facilitates domain annotation of protein data from GenBank and fasta files. For this, it harnesses the capabilities of BGClib.

Features overview

  • Annotate domains
  • Generate high quality SVG files

Additionally, if you are working with biosynthetic gene clusters, you can:

  • Use antiSMASH results as input to label proteins as core biosynthetic proteins (CBPs)
  • Extract protein (sub)sequences of CBPs
  • Print metadata of the files per BGC and CBP (e.g. protein id, CBP content, internal CBP identifiers)

And finally, if you're working with fungal biosynthetic gene clusters in particular, you also can:

  • Classify some CBPs into more specific types using included hmm models (e.g. T1PKS -> reducing/non-reducing PKS)

Other features:

  • GenBank, fasta and binary 1 input
  • Input filters (default values tailored for antiSMASH output; based on file/BGC name)
  • Domain annotation using multiple hmm libraries.
  • Save as binary files 1

Requirements

Installation of the required libraries to make BGClib work through (mini)conda is recommended. Here's a list of what you'll need. Version in parenthesis is the one that is known to work, but newer versions should work as well.

  • biopython (1.78)
  • hmmer (3.3.2)
  • lxml (4.5.0)

Documentation

Please see the BGCtoolkit wiki for the documentation

Results overview

A quick overview of results using the example output from fungiSMASH (on Aspergillus fumigatus Af293).

BGGtoolkit can produce metadata from all the input files, such as:

BGC metadata

Metadata at BGC level (BGC Metadata):

BGC antiSMASH products Core Biosynthetic Protein content Core Biosynthetic Protein IDs Core Biosynthetic Protein Identifiers Metabolites
CM000169.1.region001 T1PKS rPKS EAL87813.1 CM000169.1.region001~L0+CDS7
CM000169.1.region002 NRPS NRPS EAL90366.1 CM000169.1.region002~L0+CDS8
CM000169.1.region003 terpene
CM000169.1.region004 betalactone NRPS-like EAL90832.1 CM000169.1.region004~L0+CDS5

CBP metadata

Metadata at CBP level (CBP Metadata):

BGC Core Biosynthetic Protein type Protein identifier Protein Id Gene Id Domain architecture
CM000169.1.region001 rPKS CM000169.1.region001~L0+CDS7 EAL87813.1 KS | KS_C | KS_Ce | AT | DH | CMeT | KR | T/ACP >
CM000169.1.region002 NRPS CM000169.1.region002~L0+CDS8 EAL90366.1 A | T/ACP | C | C | A | C | A | T/ACP | C | A | T/ACP | C | C | T/ACP | C | T/ACP >
CM000169.1.region004 NRPS-like CM000169.1.region004~L0+CDS5 EAL90832.1 DMAP_binding | A | A >
CM000169.1.region005 NRPS-like CM000169.1.region005~L0+CDS9 EAL91049.1 A | A_C >

Summary

Also, a summary of all the regions found, and their CBP compositions:

AfumigatusAf293 summary file

This collection contains
* 37 BGCs
* 0 Proteins

Core Biosynthetic Composition count (BGCs):
7       NRPS
6       NRPS-like
5       rPKS
...

SVG Figures

fungiSMASH results report a 80% similarity between region CM000175.1.region001 and neosartoricin B (cluster BGC00001144 in MIBiG). Here's a comparison of both regions (top: MIBiG entry; bottom, CM000175.1.region001):

neosartoricin

Here, colored boxes represent genomic regions that will code for predicted domains. Introns are drawn by default.

Sequences

Finally, the sequences of all detected KS domains (and their metadata) can be extracted:

>CM000170.1.region002~L0+CDS9_KS1 ProteinId:EAL94057.1 GeneId:
SKIAIIGMSGRFPEADGIEAFWDLLYKGLDVHKKVPPERWDVDAHVDLTGTKRNTSKVPYGCWINEPGLFDARFFNMSPR
EALQADPAQRLALLSAYEALEMAGFVPNSSPSTQRDRVGIFMGMTSDDYREINSGQDIDTYFIPGGNRAFTPGRINYYFK
FSGPSVSVDTACSSSLAAIHLACNAIWRNDCDTAISGGVNLLTNPDNHAGLDRGHFLSRTGNCNTFDDGADGYCRADGVG
TIVLKRLEDA
>CM000169.1.region006~L0+CDS6_KS1 ProteinId:EAL91103.2 GeneId:
PFNLDRFYHPTGSHHGTTNIRQAYLLSEDVRAFDAKFFSVPPGDAEAIDPQQRLLLEVTYEALESSGHTLADLSNSNTGA
FVGLMSQDYFALNGQDVDSVPTYAASGTAASNASSRLSYFFNWHGPSMAIDTACSSNLVAVNEAVQALRNGTSRVAVACG
TNLCLSAFTFITLSKLSMLSPTSRCHMWDADADGYARGEGVACVVLKTLSDA
>CM000171.1.region002~L0+CDS2_KS1 ProteinId:EAL86536.1 GeneId:
PIAVVGMGMRLPGGVRTVDDFWDALISQKDCSSEVPQTRYNIDAFYHPDKPQSVRTRRGYFLEDDCLQKADTNFLQWIPG
FSTSELDPQQRLLLEVIWECMENAGQTGWRGKDIGCYVGVFGEDWHELTAKESQMIPRTHAFANGGFALSNRVSFEFDLK
GPSLTIATACSSSLSALHEACQALQTGSCSSAIVAGTNMLLTPSMSVTMSENMVLSPDGLCKTFDADANGYARGEAVNAV
YIKTLDKA
...

This can be useful for making phylogenetic studies of the CBPs (sub)sequences, which can include characterized data (e.g. from MIBiG) in order to to study evolution and de-replicate BGCs for prioritization.

Footnotes

  1. Currently uses Python's Pickle module for serialization of BGC or BGCCollection objects from BGClib 2

About

Methods and classes to support Biosynthetic Gene Cluster analysis

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages